[MSE] Thoughts on Bug 18575 for TPAC discussion

Hi,

I had an action item in the last MSE call to provide my thoughts on Bug
18575 <https://www.w3.org/Bugs/Public/show_bug.cgi?id=18575> (Section 2.
Source Buffer Model should be non-normative) to facilitate a discussion of
this bug at TPAC.

Here is my current thinking on the various pieces of Section 2.

1. I agree that some sections like 2.1, 2.2, 2.6, 2.7, 2.11 are redundant
with the algorithms later in the document and could probably be removed.
 Some text from 2.11 could be turned into non-normative notes in the
timestampOffset related areas of the append() algorithm.

2. Section 2.3 feels normative to me, but it may need to be reworded or
reviewed for redundant info that is already enforced by the algorithms.

3. Sections 2.4 & 2.5 probably should be moved into Section 8 and vetted
for non-normative statements.

4. Section 2.6 is covered in the append() algorithm, but the relevant steps
should be factored out into a sub-algorithm of append(). I think this will
remove the need for Section 2.6.

5. Sections 2.8(.x) should probably be marked non-normative for now and the
append() algorithm needs to have normative sub-algorithms defined to handle
all of these overlap cases. There are a number of things we have
implemented in Chrome for overlaps that aren't remotely covered by these
simple descriptions.

6. Section 2.10 feels non-normative to me since it is just saying that the
UA can remove segments if it wants to at any time.

These are just my initial suggestions for how to tackle this bug. Comments
are definitely welcome. I think this initial breakdown could serve as rough
agenda for the discussion of this bug at TPAC.

Aaron